Learn · Efficiency ratings
SEER2, HSPF2 & COP at 5°F, explained
Every heat pump and AC spec sheet is covered in these three ratings, and most shopping sites use them interchangeably. They're not the same measurement. One predicts your summer bill, one predicts your winter bill, and one tells you something neither of the others can: whether the system can still heat your house on the coldest night of the year. Here's what each one actually measures, why a "SEER2" number isn't the same as an old "SEER" number even on identical equipment, and where our own certified corpus lands on each scale.
The short answer
- SEER2 — cooling efficiency averaged across a whole summer. The number that predicts your AC or heat pump's cooling bill.
- HSPF2 — heating efficiency averaged across a whole heating season. The number that predicts a heat pump's winter bill on a typical mix of mild and cold days.
- COP at 5°F — how much heat the unit delivers per unit of electricity it draws at one specific, brutal condition: outdoor air at 5°F. The number that tells you whether it still works on the coldest night, not just on an average one.
What's good: the DOE sets a regional minimum SEER2 floor for new central AC — 13.4 in the North, 14.3 (or 13.8 for larger systems) in the Southeast and Southwest. In our own corpus of 760 certified central AC units, the median is SEER2 14.3 and the top 10% clear SEER2 16.0. For heat pumps, our 12,239-unit corpus splits by equipment type — ducted systems median SEER2 16.0, ductless mini-splits median SEER2 21.7 — while HSPF2 medians at 9.5 (top decile 11.0) and COP at 5°F medians at 2.00 (top decile 2.26).
SEER vs. SEER2 — why the 2023 change matters
On January 1, 2023, the DOE replaced the old SEER test with a new procedure called M1. The core change: M1 tests equipment against five times more external static pressure (0.5 inches of water column, up from 0.1) — a much closer stand-in for the real ductwork a system actually has to push air through, instead of a nearly unrestricted lab setup. The equipment itself didn't get less efficient; the measurement got more honest about how it performs once installed.
That means SEER2 numbers read lower than the SEER numbers the same equipment would have scored under the old test. The industry's own conversion example: a system that tested at 14.0 SEER now tests at 13.4 SEER2 — about 4% lower for the identical machine, with the exact gap varying by unit rather than a single fixed multiplier. This is exactly why a "16 SEER" figure sitting on an older blog post or an un-updated manufacturer page can't be compared straight across to a "15.2 SEER2" listing today — they were measured under different rules. Compare SEER2 to SEER2, HSPF2 to HSPF2, never a legacy number to a current one.
What these numbers don't tell you
SEER2, HSPF2 and COP are all certified lab measurements of the equipment alone. None of them capture what actually determines your comfort and your real bill:
- Install quality. A correctly charged refrigerant line, sealed ductwork and a properly matched indoor coil can matter more than a point or two of rated efficiency — and a bad install can cost you every bit of the efficiency you paid for.
- Right-sizing (Manual J). An oversized system short-cycles — it never runs long enough to dehumidify or reach steady efficiency — while an undersized one runs constantly and still can't keep up on extreme days. Neither mistake shows up in the SEER2/HSPF2 number; both come from skipping a proper load calculation.
- Noise. None of these three ratings measure sound. A high-efficiency variable-speed compressor is usually quieter, but decibel rating is a separate spec — check it separately if noise matters to you.
- Real-world weather and duty cycle. HSPF2 and SEER2 are seasonal averages built from standardized climate assumptions; your actual winter or summer, your thermostat habits and your home's insulation will all move your real bill up or down from the rated number.
Metric quick-reference
| Metric | What it predicts | Certified source | Where to see it on our site |
|---|---|---|---|
| SEER2 | Seasonal cooling bill | AHRI Directory / ENERGY STAR | Best central ACs |
| HSPF2 | Seasonal heating bill (mild-to-average days) | AHRI Directory / ENERGY STAR | Best cold-climate heat pumps |
| COP at 5°F | Whether it still heats on the coldest night | AHRI Directory (heating capacity/COP test point) | Best cold-climate heat pumps, operating-cost calculator |
Corpus figures above are computed from our normalized dataset (data current as of July 12, 2026), joined on the AHRI reference number. We publish our derived layer and cite the AHRI reference; we do not republish AHRI's directory table verbatim. DOE regional minimums per AHRI's 2023 Energy Efficiency Standards summary. SEER2 test-procedure details and the 14.0 SEER / 13.4 SEER2 conversion example per seer2.com, the DOE-aligned industry resource on the M1 test change.
See heat pumps ranked by HSPF2 & COP at 5°F → See central ACs ranked by SEER2 → Run your own $/MMBtu numbers → Browse the full database →
Common questions
Is a higher SEER2 worth it?
It depends on your electricity price and how many hours a year you run the AC — but here's a real number from our corpus. The median central AC in our database is SEER2 14.3; the top 10% clear SEER2 16.0. At the EIA U.S. average electricity price ($0.1883/kWh), that's the difference between $13.17 and $11.77 to deliver one million BTU of cooling — about $1.40 less per MMBtu, roughly 11% cheaper to run, for a system that usually also costs more upfront. Run your own numbers, including the equipment price difference, on our operating-cost calculator.
What SEER2 do I need in my state?
The U.S. Department of Energy sets a different regional minimum for split-system central AC, effective since January 1, 2023. In the North, the floor is 13.4 SEER2. In the Southeast, it's 14.3 SEER2 for systems under 45,000 BTU/h and 13.8 SEER2 at or above that size. In the Southwest, it's the same SEER2 floor as the Southeast, plus a minimum EER2 (11.7 or 11.2 by size) so equipment stays efficient specifically on the hottest days. Anything sold new has to clear your region's floor; going higher only lowers your bill further, for a higher equipment cost.
How do I convert SEER to SEER2?
There's no single fixed multiplier, because the SEER2 test procedure changed how the equipment is tested, not just the label — but the industry's own conversion example shows the scale: a unit that tested at 14.0 SEER under the old procedure tests at 13.4 SEER2 under the new one, about 4% lower for the identical machine. As a rough rule of thumb, expect a SEER2 number to read roughly 4-5% below what the same equipment would have scored as plain SEER. That means a "16 SEER" claim on an older page and a "15.2 SEER2" listing today can describe the same real-world unit — always compare SEER2 to SEER2, never SEER2 to old SEER.
What COP at 5°F is good for a cold-climate heat pump?
COP at 5°F is the number of units of heat a heat pump delivers per unit of electricity it draws with outdoor air at 5°F — a real stress test, not a seasonal average. In our corpus of 11,972 certified heat pumps with a published COP at 5°F, the median is 2.00 and the top 10% clear 2.26. Below about 1.5, the heat pump is barely beating electric-resistance backup heat on the coldest nights; above 2.5 it's still moving real heat, not just making it. See our ranked cold-climate heat pump list for the certified leaders.
Why do SEER2 numbers vary so much between ducted and ductless heat pumps?
Ductless mini-splits test inherently more efficient than ducted (central, whole-home) systems, because they skip the energy losses of moving air through ductwork. In our corpus, ducted heat pumps have a median SEER2 of 16.0; ductless mini-splits have a median SEER2 of 21.7. Comparing a mini-split's SEER2 straight against a ducted system's isn't apples-to-apples for that reason alone — match the comparison to the equipment type you're actually buying.
Written by Erin Rose. Efficiency figures trace to AHRI/ENERGY STAR certified data; computed fields ($/MMBtu, corpus percentiles) are our own derived layer, formulas in our methodology. No fabricated specs, no star ratings.